Isolation and Characterization of a Hepatoma ... - Cancer Research

More documents

Recommendations

Info

[CANCER RESEARCH 49, 6493-6497, December I, 1989] Isolation and Characterization of a Hepatoma-associated Abnormal (Des-7carboxy)prothrombin1 Howard A. Liebman William B. Cosile llematology Research Laboratory, Division of Hematology-Oncology, Boston City Hospital, and Boston University School of Medicine, Boston, Massachusetts 02118 ABSTRACT Hepatoma-associated abnormal (des-7-carboxy)prothrombin (HAPT) is a newly described tumor marker for hepatocellular carcinoma. 11Al' I has been measured in the blood of patients with hepatoma by immunoassay but has not been isolated or characterized. This paper describes the quantitative isolation and structural characterization of HAPT. Purified HAPT has the same molecular weight, amino-terminal sequence, and amino acid analysis (exclusive of 7-carboxyglutamic acid) as native prothrombin and abnormal prothrombin isolated from the blood of pa tients taking sodium warfarin. 11API' is heterogeneous in 7-carboxyglu tamic acid (Gla) content with an average of 5 Gla residues/molecule compared to 10 Cla residues for native prothrombin and 2 Gla residues for abnormal prothrombin. HUM is glycosylated in a manner equivalent to that for native prothrombin when evaluated by a concanavalin Abinding assay. These studies find structural identity between 11\ P'l and abnormal prothrombin. Therefore the findings support the hypothesis that HAPT results from an acquired defect in the posttranslational vitamin K-dependent carboxylation of the prothrombin precursor and not an intrinsic defect in the prothrombin precursor molecule. INTRODUCTION Prothrombin is the major vitamin K-dependent blood coag ulation protein synthesized by the liver (1). In the presence of sufficient quantities of vitamin K, 10 NH2-terminal glutamic acid residues undergo a posttranslational 7-carboxylation (2- 4). The resulting Gla2 residues confer metal-binding properties essential for functional activity (5). In the absence of vitamin K or in the presence of vitamin K antagonists, the vitamin Kdependent carboxylase activity in the liver is inhibited and desâ€¢y-carboxyforms of prothrombin (abnormal prothrombin) are released into the blood (6, 7). The abnormal prothrombin cannot bind metal ions and is devoid of functional activity. Conformation-specific antibodies against these abnormal forms of prothrombin have been developed and can be used to quantitate the levels of abnormal prothrombin antigen in blood (8, 9). We have reported that abnormal prothrombin antigen is increased in the blood of patients with hepatocellular carcinoma (10). Abnormal prothrombin antigen in these patients is not due to vitamin K deficiency since it did not disappear with parenteral vitamin K. The abnormal prothrombin antigen was eliminated or reduced with tumor resection or with chemother apy supporting the primary role of the hepatoma in the synthe sis of this antigen. These findings suggest that abnormal pro thrombin antigen may be a useful new serum marker of primary hepatocellular carcinoma. Subsequently, other investigators have confirmed these findings using different assay systems (11-13). Received 5/2/89; revised 8/28/89: accepted 9/6/89. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked adverlisenn'nÃ¬in accordance with 18 U.S.C. Section 17.14solely to indicate this fact. 1This work was supported by grants from the American Cancer Society (PDT- 256) and the NIH (I R29 HL .ÃŒ9C65). J The abbreviations used are: Gla. vcarboxyglutamicacid; NPT. human name prothrombin; APT. abnormal prothrombin; TBS. 0.05 M Tris-0.15 M NaCI. pH 7.4; HAPT. hepatoma-associated abnormal (des-vcarboxy) prothrombin: HPLC. high performance liquid Chromatograph)-. 6493 While the abnormal prothrombin antigen in these patients is assumed to be similar to abnormal prothrombin in the plasma of patients treated with warfarin, the hepatoma-associated ab normal prothrombin antigen has not been purified from the blood of patients with hepatoma to demonstrate a structural relationship. I have now quantitatively isolated and character ized a hepatoma-associated abnormal prothrombin from the malignant ascites of a patient. Analysis of this protein demon strates structural identity with the partially carboxylated non functional prothrombin species isolated from the blood of pa tients taking sodium warfarin. These studies further support the hypothesis that abnormal prothrombin in the blood of patients with hepatocellular carcinoma is secondary to a defect in the vitamin K-dependent carboxylation system in the pa tient's tumor. MATERIALS AND METHODS Protein Preparation. NPT was purified by barium citrate absorption, DEAE-Sephacel chromatography, and dcxtran sulfate chromatography using standard methods (14. 15). In later experiments, human native prothrombin was isolated directly from plasma on a column of antiprothrombin:Ca(Il) antibodies by modification of methods published previously (16, 17). APT was prepared by the method of Blanchard et al. (9). HAPT was purified from patient ascites fluid by ion-exchange chro matography and immunoaffinity chromatography (17). Approximately 500 ml of ascites fluid collected in 3 mM EDTA was made 1 m.M diisopropylphosphofluoride and 1 mM benzamidine. The ascites fluid was then dialyzed at 4Â°Covernight against 0.1 M potassium phosphate, pH 7.5, containing 1 mM benzamidine. Prothrombin species were purified by ion-exchange chromatography using a DEAE-Sephacel A- 50 column (3x8 cm) equilibrated with 0.1 M potassium phosphate, pH 7.5. containing 1 mM benzamidine. Dialyzed ascites fluid, after centrifugation. was applied to the column, and the column was washed exhaustively with the equilibration buffer. Bound protein was eluted with 0.5 M potassium phosphate. pH 7.5, containing 1 mM benzami dine. The eluate was pooled, made 1 mM in diisopropylphosphofluoride. and dialyzed at 4Â°Covernight against 0.1 M boric acid-1 M NaCI-1 mM benzamidine-0.1% Tween 20, pH 8.5. The prothrombin species were purified by immunoafTinity chroma tography. The eluate from the DEAE column was applied to an antiprothrombin antibody-Sepharose column (1x5 cm) and the column was washed with 0.1 M boric acid-1 M NaCl-1 mM benzamidine-0.1% Tween 20. pH 8.5. Bound prothrombin was eluted with 4 M guanidine- HC1 and then immediately pooled and dialyzed at 4Â°Cagainst 0.05 M Tris-0.5 M NaCI-1 HIMbenzamidine-HCl-0.01% Tween 20, pH 7.5. An anti-prothrombin:Ca(II) antibody-Sepharose affinity column was then used to remove contaminating fully carboxylated native prothrombin (16, 17). After dialysis, the prothrombin species were made 5 mM CaC12 and applied to the anti-prothrombin:Ca(II) antibody-Sepharose column equilibrated with TBS-5 mM CaCI2. The HAPT did not bind to the column and was collected in the flow-through. The small quan tities of contaminating native prothrombin were eluted from the column with 10 mM EDTA. Immunoassays and Coagulation Assays. Total prothrombin antigen, fully carboxylated native prothrombin antigen, and abnormal pro thrombin antigen were measured by radioimmunoassays as described previously (8, 9). The calcium-dependent prothrombin coagulant activ- Downloaded from cancerres.aacrjournals.org on July 30, 2013. © 1989 American Association for Cancer Research.
CHARACTERIZATION OF A HEPATOMA ABNORMAL PROTHROMBIN ity was assayed by measuring the acceleration of the clotting of prothrombin-deficient plasma using Stypven to initiate coagulation (9). The activation of prothrombin and HAPT to thrombin in the absence of calcium was performed using the Echis carinÃ³las venom assay (7). Protein concentrations of prothrombin and HAPT were determined spectrophotometrically at 280 nm using an E^,inmâ€žmof 14.4 (16). Analytical Electrophoresis. Polyacrylamide gel electrophoresis in the presence of dodecyl sulfate was performed on prothrombin, abnormal prothrombin, and HAPT using 10% acrylamide gels (18). Proteins were reduced using 2-mercaptoethanol. Electrophoretic analysis of the het erogeneity of the partially carboxylated prothrombin species in the HAPT was performed using 1251 radiolabeled proteins and disc gel electrophoresis (19) in the presence of 8 mM CaC12 as described by Borowski (17). Native prothrombin, abnormal prothrombin, and HAPT were radiolabeled with Nal251 by the lactoperoxidase method (20). Electrophoresis was performed on the tube gels with a 7.6% acrylamide, 0.3% bisacrylamide resolving gel and a 1.8% acrylamide- 0.22% bisacrylamide stacking gel. Gels, buffers, and samples were made 8 mM in CaCl2. After electrophoresis the gels were sliced into 1-mm fragments and counted for 1251 in a Beckman Gamma 8000 scintilla tion spectrometer. Ammo-terminal Sequence and Amino Acid Analysis. The amino terminal sequence of HAPT was determined by automated Edman degradation using a Beckman 8900 spinning cut sequenator equipped with a cold trap. The phenylthiohydantoin derivatives were analyzed on a Waters HPLC using a Waters C18-/iBondapak column with an ethanol gradient in ammonium acetate, pH 5.1 (21). Amino acid analysis was performed on a Beckman model 119CL amino acid analyzer equipped with a Beckman model 126 data system. Proteins were hydrolyzed in 6 MHC1 at 110Â°Cfor 24 h. For quantitation of Gla, the proteins were hydrolyzed in 2 M KOH for 22 h at 110Â°C (17, 22). The Gla composition was quantitated after alkaline hydrolysis by HPLC using a fluorescence detection procedure (17, 22). Lectin Blotting Studies. A comparison of the level of glycosylation of HAPT and NPT was performed by the binding of 1251-labeled concanavalin A to these proteins immobilized on nitrocellulose paper. Serial dilutions of HAPT and NPT in TBS were applied to the nitro cellulose paper. The paper was then incubated for l h in 3% bovine serum albumin in TBS. After the blot was washed 4 times with TBS, it was incubated for 3 h at 37Â°Cwith '25I-labeled concanavalin A in 3% bovine serum albumin-TBS. The blot was then washed 5 times in TBS, dried, and exposed to Kodak X-O mat R Film for 2-6 days. Quantita tion of the binding of the concanavalin A was performed by densitometric scanning of the autoradiographs using an EDC densitometer (Helena Laboratories). RESULTS Purification of HAPT. HAPT was purified from the ascites fluid of a patient with widely disseminated hepatocellular car cinoma. Approximately 2.8 liters of ascitic fluid were used for the isolation of HAPT. The patient had markedly elevated blood levels of HAPT and had previously been reported (Ref. 10; Table 2, Patient 3). Listed in Table 1 are the concentrations of NPT and HAPT in the blood and ascites fluid of this patient as measured by radioimmunoassay. The ascites fluid concentra tion of HAPT was equivalent to the levels found in the patient's blood, while NPT was only 35% of the blood levels. After purification, 4.2 mg of HAPT were obtained from the ascites fluid. This was a final preparative yield of 38% of the estimated total concentration of HAPT in the ascites fluid (Table 1). The concentrations of total prothrombin antigen, native pro thrombin antigen, and abnormal prothrombin antigen were measured in the purified HAPT preparation by competition radioimmunoassays (Table 2). The concentrations of these an tigens were compared with the estimated concentration of HAPT as determined spectroscopically. Approximately 92% of the purified HAPT competed with NPT in the radioimmuno- Table 1 Purification of hepawma-associated abnormal prothrombin AmountSerum (//g/ml") NPT (Kg/ml0) Ug/ml")Ascites HAPT NPT (Mg/nila) HAPT (â€žg/ml") (liters)Total Volume HAPT (mg) Isolated HAPT (mg*) HAPT yield ("Â¿)603.921 Â°Determined by competition radioimmunoassay. * Determined spectroscopically. 3.8810.864.2 Table 2 Immunochemical characterisation of purified HAPT HAPT" TPT*NPT* APT*Concentration (jig/ml)640 588 11 511% 38 ofHAPT100 92 2 80 *Concentration determined spectroscopically. *Concentration determined by competition radioimmunoassay. assay for total prothrombin antigen. The native prothrombin antigen was measured in the HAPT preparation using the antiprothrombin:Ca(II) antibodies. Anti-prothrombin:Ca(ll) anti bodies bind only fully carboxylated native prothrombin species capable of undergoing a metal-induced conformational transi tion. Less than 2% of the purified HAPT competed with NPT in the radioimmunoassay for native prothrombin antigen. Antiabnormal prothrombin antibodies bind abnormal prothrombin but not fully carboxylated native prothrombin (9). In the com petition radioimmunoassay for abnormal prothrombin antigen, 80% of the HAPT competed with APT. Approximately 10% of HAPT as measured by the immunoassay for total prothrom bin antigen failed to bind to either the anti-prothrombin:Ca(II) or anti-abnormal prothrombin antibodies suggesting the pres ence of a population of partially carboxylated prothrombin species (17). Structural Characterization of HAPT. The isolated HAPT was studied by dodecyl sulfate-gel electrophoresis. The protein appeared >95% pure and migrated as a single band in the presence of 2-mercaptoethanol with the same mobility as NPT (Fig. 1). The amino-terminal sequence of the HAPT was com pared to that of NPT in order to determine whether the varia tion in Gla content could be attributed to differences in the primary structure of this region. The first 14 amino acid resi dues of the HAPT were identical to those of NPT and the published known sequence of prothrombin. These results indi cate that there are no differences in the amino-terminal se quences of HAPT and NPT between residues 1 and 14, nor does HAPT possess an unprocessed leader sequence similar to Factor IX Cambridge (23). However, I cannot exclude the possibility of a mutation involving other Gla residues at posi tions 16, 19, 20, 25, 26, and 32 of the prothrombin aminoterminal region. Specific identification of Gla content in NPT and HAPT cannot be obtained by this technique since the formation of the phenylthiohydantoin derivative of Gla results in its conversion into phenylthiohydantoylglutamic acid. The amino acid composition of the acid hydrolysate of HAPT showed no significant difference from NPT. The Gla compo sition was quantitated after alkaline hydrolysis by HPLC and fluorescence detection. The Gla content for NPT, APT, and HAPT is presented in Table 3. These results indicate that 6494 Downloaded from cancerres.aacrjournals.org on July 30, 2013. © 1989 American Association for Cancer Research.
Page 1: Isolation and Characterization of a
Page 5 and 6: NPT HAPT 10 5 2.5 1.25 CHARACTERIZA

Isolation and Characterization of a Hepatoma ... - Cancer Research

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?